Tuberculosis (TB) is an infectious disease with high incidence and mortality rates worldwide (Chakaya et al., 2022). It is estimated that there were 10.8 million new cases of TB worldwide in 2023, with an incidence rate of 134 cases per 100,000 individuals. Although the lungs are the most common site of TB infection, TB can affect almost any part of the body (Trajman et al., 2025). Whereas pulmonary TB accounts for most of the TB cases worldwide, there has been an increasing incidence of extrapulmonary TB (EPTB) in both high- and low-income countries since the mid-1980s, with certain groups, including immunocompromised individuals and young children, exhibiting a higher frequency of EPTB (Kohli et al., 2021). EPTB accounted for 15% of the 6.3 million TB cases recorded worldwide in 2016, with the prevalence ranging from 8% in the Western Pacific region to 24% in the Eastern Mediterranean region. In England in 2019, 59.9% of TB cases were attributed to EPTB (with or without disease at another site). Additionally, 33.1% of all TB cases are categorized as TB lymphadenitis (TBLA) (Trajman et al., 2025). More than one-quarter of pulmonary TB patients also exhibit extrapulmonary manifestations (Lucey et al., 2022). The most common site of extrapulmonary TB is the lymph nodes, including the intrathoracic lymph nodes (Armstrong et al., 2019). EPTB, particularly mediastinal tuberculous lymphadenitis, presents additional diagnostic challenges relative to pulmonary TB. This result is primarily due to the need for invasive sampling procedures, which necessitate specialized expertise and, in certain instances, may involve a degree of risk to the patient. Furthermore, owing to the paucibacillary characteristics of EPTB, achieving diagnostic confirmation using the existing microbiological methods can be challenging (Chakaya et al., 2022). Patients with mediastinal lymphadenopathy require biopsy sampling to provide a pathological and/or microbiological diagnosis, particularly when this is an isolated finding (Lucey et al., 2022).

Endobronchial ultrasound-guided transbronchial needle aspiration (EBUS-TBNA) is a minimally invasive intervention that has been demonstrated to provide a superior diagnostic yield compared with conventional TBNA and shows results comparable to those of mediastinoscopy for lymph node staging in patients with primary lung cancer (Rodriguez et al., 2024). EBUS-TBNA is recommended by BTS guidelines as a safe and effective technique for the assessment of hilar and mediastinal lymph nodes in patients with suspected lung cancer and sarcoidosis (Du Rand et al., 2011). The indications for EBUS-TBNA have been expanded beyond malignant diseases to include benign diseases, such as sarcoidosis, tuberculous lymphadenitis and fungal infections (Eom et al., 2015; Berger et al., 2016). There is increasing evidence supporting the diagnostic utility and safety of EBUS-TBNA in the diagnosis of intrathoracic TBLA, particularly in patients with a high preprocedure clinical suspicion of TB (Chhajed et al., 2019).

The needle rinse fluid and biopsy tissue obtained from EBUS-TBNA can be used to detect pathogenic microorganisms, and molecular biology methods can be used to detect Mycobacterium tuberculosis (MTB). Consequently, this study aimed to evaluate the diagnostic utility of flushing fluid and biopsy tissue obtained through EBUS-TBNA in the identification of mediastinal lymph node TB.

The primary aim of this study was to evaluate the diagnostic significance of EBUS-TBNA fluid in detecting intrathoracic lymph node TB. The patient cohort mainly comprised young males with minimal comorbidities, indicating no clear demographic predisposition for the development of intrathoracic lymph node TB. Previous literature has reported that mediastinal lymph node TB represents a common form of TB, accounting for approximately 8% to 10% of TB cases (Baran et al., 1996). Consistent with these findings, this study revealed that intrathoracic lymph node TB frequently co-occurred with pulmonary TB, and isolated cases of intrathoracic lymph node TB were relatively infrequent.

Intrathoracic lymph node TB lacks distinctive clinical symptoms, and abnormal chest CT images may be its unique manifestation; it often simulates malignancy or other inflammatory diseases (Navani et al., 2011; Nin et al., 2016). In this study, differences were observed only in respiratory symptoms between patients with isolated intrathoracic lymph node TB and those with pulmonary TB. No significant differences were found in TB-related systemic symptoms, inflammatory markers, nutritional status, white blood cell counts, neutrophil and lymphocyte levels, or lymphocyte subsets. These findings suggest that the clinical presentation of intrathoracic lymph node TB is nonspecific, corroborating the findings of previous reports.

Owing to its hidden location, the diagnosis of intrathoracic tuberculous lymphadenitis is usually challenging because of the difficulty of the approach and the lack of specific clinical and radiological characteristics. The acquisition of intrathoracic lymph node specimens and the use of sensitive MTB detection methods are two primary obstacles to diagnosis. Ye et al. A meta-analysis of the diagnostic efficacy of EBUS-TBNA in intrathoracic LNTB revealed a pooled sensitivity of 0.80 (95% CI, 0.74–0.85) and a specificity of 1.0 (95% CI, 0.99–1.00) on the basis of 7 studies with sufficient data for these calculations. Another London-based study by Navani et al. reported that the sensitivity of EBUS-TBNA was 94% (95% CI, 0.88–0.97). Olivia Lucey et al. confirmed the utility of EBUS-TBNA in the diagnosis of intrathoracic TB in an undifferentiated cohort of patients with mediastinal lymphadenopathy of unknown etiology (Sun et al., 2021).

Although a consensus among experts has established that the identification of characteristic pathological lesions, namely, chronic granulomatous inflammation with caseous necrosis, via histopathological examination serves as a diagnostic marker for lymph node TB (Sun et al., 2021), in the absence of direct evidence of MTB and when only nonspecific pathological features such as inflammatory exudation, cellular proliferation, or necrosis are present, histopathology often yields suggestive rather than definitive diagnoses (Xu et al., 2024) Given that TB is an infectious disease, etiological confirmation remains fundamental for diagnosing intrathoracic lymph node TB. Culture-based strain identification and drug susceptibility testing are regarded as the diagnostic gold standards; however, these procedures may take several weeks to months to yield results (Demers et al., 2016). Consequently, early and rapid diagnostic methods are urgently needed. Molecular biological techniques for detecting MTB fulfill this requirement. The Xpert MTB/RIF assay, a molecular diagnostic tool endorsed by the World Health Organization (WHO), enables the detection of MTB and rifampicin resistance in pulmonary specimens (e.g., sputum) within approximately two hours (Haridas et al., 2022). Furthermore, the WHO recommends the application of Xpert MTB/RIF for nonpulmonary specimens, such as cerebrospinal fluid and lymph node samples, in cases suspected of extrapulmonary TB (Maynard-Smith et al., 2014). The MTB nucleic acid (TB-DNA) detection kit, developed in China, relies on the use of dual real-time polymerase chain reaction (PCR) technology combined with TaqMan probe technology to detect mycobacterial nucleic acids, completing analysis within three hours and facilitating rapid diagnosis. This PCR-based assay has received certification from the Chinese Food and Drug Administration and enables prompt detection of MTB nucleic acid. Its diagnostic performance is comparable to that of acid-fast bacilli detection in sputum smears, assisting in the diagnosis of smear-positive pulmonary TB and enhancing diagnostic accuracy in smear-negative cases. It represents a rapid and sensitive method for pathogen identification (Cheng et al., 2024). A systematic review and meta-analysis demonstrated that the Xpert MTB/RIF assay has high diagnostic accuracy for detecting lymph node aspirates, although the majority of the included studies focused on peripheral lymph nodes (Kohli et al., 2021). Another study within this meta-analysis evaluated the diagnostic performance of Xpert MTB/RIF on samples obtained via EBUS-TBNA and concluded that this technology has satisfactory sensitivity for diagnosing mediastinal tuberculous lymphadenitis (Mondoni et al., 2023).

A total of 156 lymph node stations were sampled from 63 patients. The total number of sampled stations exceeded the number of patients because multiple stations (ranging from 1 to 5) were punctured per patient. The most frequently sampled stations were the subcarinal (station 7; 82.5%) and the right lower paratracheal (station 4R; 66.7%). Given that the majority of patients undergo EBUS-TBNA for cancer evaluation, it is critical to acquire an adequate quantity of core tissue to facilitate comprehensive histopathological analysis and molecular diagnostics. Specimens obtained via EBUS-TBNA can be subjected to diverse analytical techniques, including cytological assessment, flow cytometry, and cell block preparation for immunohistochemical or molecular investigations, as well as microbiological evaluations such as culture (Ko et al., 2020). Prior research focusing on TB mainly involve the use of biopsy samples acquired through EBUS-TBNA for the detection of MTB, with limited studies employing EBUS-TBNA needle rinse fluid for this purpose (Ko et al., 2019). The EBUS-TBNA needle rinse fluid, obtained by irrigating the needle lumen with sterile normal saline after core tissue retrieval, represents an additional sample that can be rapidly collected within seconds. Investigations have demonstrated that the MTB burden in EBUS-TBNA flushing fluid is relatively low; specifically, only 1.9% of patients with intrathoracic TB lymphadenitis yielded positive smear results from this fluid. Nonetheless, routine mycobacterial culture of EBUS-TBNA needle flushing fluid has proven effective for TB detection and monitoring [30]. Given the comparable sensitivities of the Xpert MTB/RIF assay and mycobacterial culture, it is reasonable to infer that molecular diagnostic modalities such as Xpert MTB/RIF are well suited for detecting MTB in EBUS-TBNA needle flushing specimens.

Despite the retrospective nature of this study, our findings indicate that both the Xpert MTB/RIF assay and the TB-DNA method yielded significantly higher positivity rates in EBUS-TBNA flushing fluid than in EBUS-TBNA biopsy tissue samples. These results underscore the considerable diagnostic value of EBUS-TBNA flushing fluid in detecting intrathoracic lymph node TB. Furthermore, the analysis of flushing fluid offers additional benefits, as it does not necessitate extra puncture procedures or increased operation time.

However, retrospective design inherently suffers from potential selection and information bias. Since the diagnostic value of needle rinse fluid for TB remains unclear, not all patients underwent concurrent collection of biopsy specimens and rinse fluid for testing. In clinical practice, due to considerations of patients’ medical costs, needle rinse fluid was not collected for tuberculosis testing in patients with non-TB lymph node disease. In addition, this study has several other limitations. First, the retrospective design may introduce selection bias, potentially influencing the findings. Therefore, future prospective cohort studies are warranted to validate these findings. Second, among the 63 patients diagnosed with intrathoracic lymph node TB, not all underwent collection of both EBUS-TBNA flushing fluid and tissue samples for Xpert MTB/RIF and TB-DNA testing, primarily owing to sample availability and financial constraints, resulting in a reduced sample size for analysis. Although this study had a relatively small sample size, a statistically significant difference in the positive rate was observed in the Xpert MTB/RIF group (P = 0.004), suggesting a substantial effect size and a certain degree of reliability in the findings. Third, the limited number of MTB cultures performed, which is attributable to sample quantity and economic factors, precluded the analysis of culture results from the EBUS-TBNA flushing fluid and tissue specimens.

In summary, we recommend that EBUS-TBNA flushing fluid should be used as a diagnostic specimen for intrathoracic lymph node TB. Furthermore, when resources permit, simultaneous collection of both EBUS-TBNA flushing fluid and tissue samples for MTB detection is recommended to increase diagnostic accuracy.